Running dry?

In article ,
Greg Copeland wrote:

In September 2004 issue of AOPA Flight Training, Mark Cook has an article,
"No Fueln' Around". Under the "Selector boy" side article, he mentions
that he runs some of his tanks dry in his Bellanca Viking. In at least
one of John Deakin's articles
(http://www.avweb.com/news/columns/182044-1.html), he not only recommends
running tanks dry but puts forth a powerful argument that it's a
responsible fuel management strategy. Furthermore, Deakin also offers
that he has never found an NTSB accident report related to a failed engine
start when running a tank dry and switching to the next. Both guys
recommend setting a timer a couple of minutes before the tank should run
dry; which acts of both early warning and as validation of your
anticipated fuel consumption.

Is this common? How many run their tank(s) dry as part of their fuel
management strategy? If you don't run dry, why not? Aside from the heat
beat skipping which is sure to follow the first couple of times, what's
the down side to this strategy?

I was taught to run aux tanks dry, as a matter of fuel management. This
technique is best on carbureted engines, as restart is just about
instantaneous, as soon as th float bowl fills. On fuel-injected engines,
it takes a few seconds (which seems like hours) to get fuel to the
engine and back running.

The philosophy is that it is best to end a flight with all of yoy=ur
available fuel in a single tank, to prevent starvation at critical
times. On the old, pressure-carburetor Bonanzas, the fuel return fed
back to the left main tank only (about 2-3 gph). The procedure was to
run that tank dry, switch to the aux tank(s), run dry, switch to right
main an run it dry. You are left with an hour's worth of fuel in the
left main and no longer have to switch tanks for the duration of the
flight.

You can catch the "tank dry" point by monitoring the fuel pressure gage
and switch as soon as you see it flicker.